Analysis of Child Deprivations and Well-being Inequalities in Kenya

Abstract

Unidimensional approaches, especially those based on income or consumption and expenditure, previously dominated evaluation of the status of household or individual well-being. This thesis argues that the use of several non-metric indicators would be more suitable since well-being is a multidimensional phenomenon. Thus, measurement of well-being in terms of income or consumption and expenditure does not capture fully deprivations experienced by children because they are not economic agents. Furthermore, discussions based on income or consumption and expenditure indicators alone are often insufficient for the formulation of policies aimed at reducing deprivations in non-monetary dimensions of well-being. Finally, besides income, individuals and households also differ in health status, nutritional status, educational attainment, and housing conditions among others, which should be considered during evaluation of their well-being. There is little documentation on the study of child deprivations and well-being inequality from a multidimensional perspective in Kenya. This thesis reports the use of a multidimensional perspective in the measurement of deprivations and well-being inequalities among Kenyan children. Using nationally representative and comparable data from the Kenya Demographic and Health Surveys (KDHS) for the years 1993 to 2014, the thesis examined child poverty and well-being disparities in Kenya. The investigation of multidimensional child well-being disparities followed an evaluation of the levels and patterns of different child deprivations. The study looked into the correlations between different child deprivations as well. The deprivations in seven child-specific dimensions of well-being—nutrition, health, education, water, shelter, sanitation, and information—are specifically reported in this thesis. The outcomes were produced using the Bristol Deprivation Approach and the Alkire and Foster (AF) Methodology, two frontier methodologies from the literature. The study's methodological framework has three components. First, the prevalence, depth, and severity of multiple child deprivations were examined using the Bristol framework and the Alkire and Foster methodology. Second, composite well-being indices were calculated using multiple correspondence analysis (MCA) to measure multidimensional well-being inequalities among Kenyan children. This study further decomposed the overall well-being inequalities to account for xvii the contribution of various characteristics to inequality using regression-based decomposition approach. Third, we counted the dimensions in which each child experienced deprivation and ordered the children from those who had no deprivation to those who experienced deprivation in four or more categories. As a result, we examined the distribution of multiple child deprivations by various characteristics and examined parameters related to multiple child deprivations using an ordered logit regression model. The results of the first essay show that the majority of Kenyan children are deprived in the areas of information, shelter, and water, in that order. The lowest rates of deprivation were observed in health dimension followed by education dimension. However, it was noted that the rates of deprivation in education dimension increased from 2008 to 2014. Multidimensional poverty index among Kenyan children declined from 0.37 to 0.26 between 1993 and 2014. This observation was attributed mainly to reduction in multidimensional headcount ratios from 72.77 percent in 1993 to 50.76 percent in 2014. The average deprivations among the children remained constant during the study period. The highest contributors to multi-dimensional child poverty (MPI) were information, water, and shelter dimensions while the lowest contributors were health, nutrition, and education dimensions. Children from North Eastern and Eastern regions suffered the highest deprivations in all dimensions compared to other regions. Results from the second essay indicate the existence of modest and reducing well-being inequalities among the children. The Gini coefficient index was 0.29 in 1993, but declined to 0.23 in 2014. The Theil index was 0.15 in 1993 but declined to 0.09 in 2014 when the inequality aversion parameter is low, α=1. The multidimensional inequality when the inequality aversion parameter is high, α=2 showed slight dispersion as the entropy index declined from 0.16 in 1993 to 0.09 in 2014. Further, this essay found that inequality was relatively higher for girls, non-urban children, and children from female-headed households. The key contributors to multidimensional inequality were nutrition, housing, and information dimensions in that order. The most important determinants of inequality were identified as additional level of education of the mother, access to electricity, and non-urban residence of children. xviii From the third piece, numerous observations were made. First, rural children are more susceptible to numerous deprivations than their urban counterparts. Second, maternal education significantly reduced the incidence of multiple child malnutrition. Thirdly, children from female-headed homes experienced greater deprivation than their male-headed counterparts. Fourth, compared to other regions, children in the Nairobi region had the lowest deprivation rates, while those in the North Eastern region had the highest. Fifth, children from religious homes had higher outcomes than those from families with no religious affiliation. The empirical model revealed that children with higher birth orders, twins, and those living in rural areas were more likely to experience multiple deprivations, whereas children with mothers with higher levels of education and employed respondents were less likely to experience multiple deprivations. In addition, boys were more susceptible to multiple deprivations than girls, and the prevalence of multiple deprivations was higher among twins than among singletons. The findings of this thesis suggest policies for tackling multiple forms of poverty. The thesis suggests streamlining of development in rural, arid, and semi-arid lands by encouraging investment in infrastructure and public services like hospitals, schools, electricity, water supply schemes in rural areas and ASALs. Households lacking access to sanitation facilities for proper disposal of human waste should be sensitized about the importance of constructing and using toilets to prevent disease outbreaks. At the same time, there is need for enforceable regulations on the proper disposal of human waste to conserve the environment. In terms of information, the government should introduce radio programmes in schools and community cinemas to expose children to events in the world outside their immediate environment as well as improve their cognitive development. The government could consider tax exemption for broadcast equipment and removal of the requirement for licencing to make this equipment affordable for most households so that children can access information through the broadcast media. To spare children from growing up in conditions of severe hunger, the government should introduce sustainable targeted programmes for providing access to adequate food for vulnerable families with children, especially those under five years. These programmes could include relief food and expanded cash transfer programmes to vulnerable sub-groups. In the long term, this study suggests establishment of a universal childcare programme where families receive assistance in xix cash or in kind to enable them access food and education. Finally, in addition to encouraging research and development to identify innovative ways of constructing cheap and affordable houses, the government should also reduce taxes and levies on building materials such as iron sheets, cement, sand, etc. to enable households to construct better houses. This thesis makes a significant contribution to the study of child poverty. First it contributes to the conceptualization and reformulation of child well-being in a multidimensional perspective. Unlike past studies which examined child well-being based on income or consumption and expenditures, this thesis analysed child well-being using the whole gamut of child-specific dimensions. Secondly, the thesis contributes to the literature of multidimensional child well-being inequality. The innovative aspect of this thesis is that it developed a composite well-being index by using multiple correspondence analysis to investigate the distribution of children's well-being across seven child-specific variables. The index was then used to calculate multidimensional inequality. Thirdly, the thesis makes a contribution to measuring the severity of child deprivations by counting the number of domains in which a child is deprived and ranking them from a child who is not deprived to a child who is deprived in four or more domains with the assumption that a child deprived in one domain is better-off than a child deprived in two, three, four or more domains. Lastly, the thesis prescribes policy implications to reduce the child deprivations.